WiFi Tracking, Robot bees, Autonomous vehicles, jetpacks & more
|mm||Jun 3, 2019|| 3|
🌊 Welcome to the third edition of 🔮Technologically Clairvoyant🔮
No announcements this week, so let’s get right into it 🎪
Following on from last week, the Netherlands publicly admits to using WiFi tracking. However, unlike TFL, they detail more on why and how. Firstly, lets cover the security concerns. MAC addresses are uniquely identifiable addresses which can identify a network card. Typically you have one network card per device, so they can be used to identify a device. When they track your device, the MAC address is hashed, then sent to a server, where it is salted. The salt changes everyday, and is not stored on a computer. They then cut some of the hash out, so there is no way that the hashed MAC address can be traced to an individual .The hashing algorithm they use is Bcrypt. I have to admit, hashing locally in the train station is genius.
Normally, client-side hashing sucks. When you hash client-side, the hashed passwords becomes the actual password. That means you will be storing the plain text in the database (the hashed password is the plain text).
But, MAC addresses are different. We know they are entirely unique, and can be used to identify a single device. By hashing them client-side, the hashes will become the IDs of the MAC address. Meaning that on the server, there is no way to get a MAC address. You can only get hashes, which do not give information on what device you own.
There is absolutely no way to personally identify you, all while keeping the benefits of having individual devices tracked through stations. It’s clear from this technique that they have put a lot of thought into the privacy of individuals, while still maintaining the benefits.
They also admitted to using Bluetooth tracking, which was discontinued last year. If you’re interested in this, do check out the link. They detail exactly why they track people, what for, how the data is being stored and for how long. This is a typical GDPR setup. All companies have to give you this information if you ask for it, they’ve published it online. I’ll let you know if I see any GDPR inquiries of what the WiFi tracking on the tube system is for / how they do it.§
The AI has an accuracy of 31%, compared to traditional methods at 18%. Something else interesting from this article is:
“Black women are 42% more likely to die from breast cancer”
The researchers at MIT believe this is due to lack of healthcare access. They hope their AI can be implemented around the world to provide cheap screenings for women of all kind.
Plastic bottles take 400 years to naturally decompose. If our plastic bottle creation rate carries on, every 400 years (when one bottle decomposes) 12610000000000000 plastic bottles will be created. It’s estimated that there will be more plastic in the ocean than fish by 2050.
This isn’t directly caused by 5G. The frequencies that are planning to be used by 5G disrupt satellites from detecting water vapor, disrupting a part of weather forecasting (but not all weather forecasting). The FCC and NOAA are working together to make sure that companies do not use the frequencies which can effect weather forecasting. Speaking of 5G, let’s address some other 5G issues.
Can 5g kill you?
No. It can’t kill you. Can it harm you? Very, very unlikely. I’m going to list some of the things that people use as ‘proof that 5G can harm us’ and objectively look at each one.
Senate Bill 637
This is a law from the United States of America. A lot of news articles / conspiracy theorists claim it’s senators fighting to protect us from 5G, and that this law wasn’t put into place.
Turns out, this law is in place. It was ratified. But the bill isn’t about the ‘dangers’ of cellphone towers. Here, you can read it:
The bill would cap the amount the state and local governments could charge for zoning, permits and other fees imposed on these “5G” networks.
This law places a limit on how much the state and local governments can charge for permits, zoning and other fees of putting 5HG into place. It’s nothing related to the ‘dangers’ of 5G. You can read it here.
False. The original source of this is Erin Elizabeth. It was posted to her blog, Health Nut News. This same blog claims that vaccines cause autism and other strange things. What’s funny is that her website is dedicated to selling products to protect you from all these ‘harmful’ things.
Can 5G kill humans?
Absolutely not. This is absurd. Firstly, there’s a 20 - 30 year test in Europe on 30,000 people which monitors how harmful phones are to humans. This study is still ongoing, but so far they’ve found nothing. Secondly, infrastructure like this is tested intensively.
4G was announced in the late 90s / early 2000s. It was officially announced in 2002. I personally got 4g 4 or 5 years ago. That’s an awfully long time for them to test things, to make things right. A lot of networking has been announced for 20+ years and still hasn’t come into effect.
Even if 5G was dangerous, it would be fixed far, far before you can even buy the first phone that supports 5G. Now, I do recommend you read for your own. Do not trust everything you read online, including this newsletter. Find the original source, always. And test that source subjectively like I have done so here.
This YouTube channel is dedicated to creating & distributing Deepfake videos. While there’s no technological breakthroughs, I have to admit that some of the videos are fun 😂
They claim they’ve found the identities of 100,000 porn stars so far. Just to be clear, there is no proof he has done this. However, it’s caused quite a stir on Twitter and other social media sites. As a computer scientist, it is definitely possible this is real. As we saw 2 weeks ago, Facial Recognition is being used in airports for passports. It’s very possible to find someone’s social media account using multiple images of them (via video).
The user posted this on Chinese site Weibo. It is difficult for me to find the original source, as it doesn’t index for me on English Google. However, there are screenshots of the original source here.
The report claims that it’s made the cost of extracting lithium (usually £20,000 per ton) down to £2180, a 817% decrease. This can bring electric vehicle costs down. But, with the Trade war Between China And America it looks like American EV manufactures will have to wait to experience this decrease in production costs.
This article has no hard numbers to prove that autonomous vehicles are no safer, however, that is where the problem lies. Tesla is refusing to release data on its crashes. The numbers that Tesla does release show that autonomous vehicles are getting safer, but perhaps right now they are not as safe as the industry hopes them to be. While I’m pro autonomous vehicles, it’s clear that these vehicles - in their current state are not cut out to work in most places.
We had hackers in the 70s / 80s who could do all kinds of cool things. Free long distance phone calls, free television cable and more. I believe that we’re on the verge of a biohacking revolution. Millions of people with diabetes don’t have access to insulin globally. And if they do, it might cost money. Biohackers aim to create a protocol for “do-it-yourself insulin”.
This is an extremely fascinating presentation. The short term clinical goal of this group is to decrease the number of unnecessary surgeries. In their words:
“80% of high risk biopsies are benign”
We know that AI can predict cancer better than humans, but it simply does not have the data set. This team is using NLP to process text documents, reports generated by machines (such as heart rate monitors) and every single medical record. Using this massive data set, they’re going to train a neural network to predict cancer.
Looking into my crystal ball 🔮 I can imagine this being used for so much more. Imagine wearing a fitness tracking device that tracks every single thing about you. You can track food and water intake too. All of this will be combined with doctors reports and past medical data. In the background, a machine will use NLP to process all of these documents.
Using neural networks, the machine will run tests for every single disease / illness under the sun. Some diseases exist for years before their first symptoms show. Imagine how many lives will be saved if in the background this unified machine checks on you 24/7.
Let’s go one step further. I don’t want some machine to know all my private health stuff? I want it all to be encrypted. I want it so if they’re hacked, no one can read it. How do you train a neural network on things that are encrypted?
You can use fully homporphic encryption. You can perform calculations on data without knowing what the data is. E(a) * E(b) = E(a * b) - where E(x) is the encryption function being applied to x. Imagine in the future being able to train a neural network without knowing any single persons data. Of course, this is looking into my crystal ball now. As far as I know, there is no fully homorphic encryption scheme. And no one has published anything on using it in this way. But it’d be a cool thing to have in 100 years.
Andrew Birch (a one-man hacker show) created an incredibly small 3D printer and demoed it by printing Gengar. The 3D printer looks to be about the size of a foot. I can imagine a day where everyone walks around with 3D printers and prints what they need. Need a screwdriver? No worries, I’ll print one now.
Most snails are born right. That is, their shell coils to the right. In fact, almost all snails are. Left snails are often lonely.
The technology is important for understanding why 1 in 10,000 people are born with situs inverus, a condition where their internal organs are flipped like a lefty snail’s shell.
Traditionally it takes many months to manufacture, test, and deliver a conventional combustion chimer. This new technology is expected to reduce that time considerably.
Bio-printing is an additive procedure where biomatierals such as cells and growth factors are combined to create tissue-like structures that imitate natural tissues. Typically low-end bioprinters cost £10,000. While higher end ones cost £170,000. This printer can be built with £375.
Jetpacks Are Real. Just like in the movies. But, there are too many problems that outweigh the advantages.
For one, humans aren’t meant to fly. We’re not aerodynamic at all, requiring a relatively large amount of force to lift into the air. Two, the fuel used for jetpacks is the same fuel used for rockets. Combining hydrogen with an oxidizing agent. Rocket fuel isn’t very good for the environment, as we can all imagine.
Thirdly, they don’t fly very long. The upper limits seem to be around 90 seconds. The fastest speed set in a jetpack is 32.02 MPH.
Robot bees exist. They can pollinate flowers, swim underwater (for some reason) and stick to surfaces just like regular bees. Scientists have seen the mass extinction of bees and decided to just build robotic bees.
On the one hand, robot bees are cool. They can pollinate 24/7. On the other hand…. I don’t understand why scientists have decided to go straight to replacing bees instead of trying to save them. In the distant future, would it be ridiculous to think that all of our animals on earth are replaced by robots?
We can replace dogs, bees, fish, cats and all sorts. In the distant future, I can imagine the rally cry’s for replacing animals. “Bees go to sleep, our robots don’t. They do more than bees do”.
“Lions are vital in the circle of life, but they kill so many humans. Think of our robot lions which have learned not to harm humans”
Volvo reportedly use this headset to test / help build cars. What’s impressing me is the idea of using AR within cars. AI can drive better than humans, but sometimes humans can see things better than AI can. Using AR, I can imagine the windshield of the car being augmented to provide details about the environment to enable drivers to drive with the AI - as opposed to either the AI or human driving.
The tiny robot hops around, as opposed to walking or driving. What’s amazing to me is that out of all the robots I’ve seen, this robot shows the most emotion - despite only being able to hop.
Until next time,